Interactive comment on “ Applying receptor models Unmix and PMF on real data set of elements in PM for sources evaluation of the sea coastal side region ( Southeast Adriatic Sea ) ”

General Comments: In the pr sent manuscript, Dordevic et al. apply two advanced and popular receptor modelling techniques (Unmix and PMF) to a dataset of concentrations of 11 trace metals in PM samples obtained from a coastal site close to the Adriatic Sea. The dataset of metal concentrations is re-used from an earlier study (Dordevic et al., 2005), where two other receptor models (enrichment factor analysis and PCA) have been applied to the data. My major problem with the manuscript is that I do not understand what it is actually aiming at:

In addition, if the goal really was a refined or more detailed source apportionment then AMTD might not be the most suitable journal choice.
-Or is the main goal of the manuscript an evaluation of the (more complex) Unmix and PMF models as compared to the (more simple) approaches of the previous study, as is stated in P4948L15-19? Such evaluation would certainly fit better into the journal scopes, but I do not see which part of the manuscript would really critically evaluate the models. To me, the whole study seems to be a sheer application of two (further) receptor models to an existing (and published) dataset, which does not represent enough scientific significance to justify publication in AMT.
If at all, the manuscript can only be considered for publication in AMT after major revisions, which would actually include a complete re-write of several manuscript sections (abstract, discussion, conclusions).
Specific comments: P4942L2-13: The abstract is poorly written. It lacks important details on the dataset (location of the sampling site, time of sampling, size of dataset, etc.), it only lists the number of possible solutions of the models without really relating them to each other or concluding on their quality, it does not give the physical meanings for all of the mathematically suggested sources, and is in part even contradictory: "traffic is not a significant anthropogenic source at the sampling site" (L5) vs. "...more realistic solution that includes ... traffic as dominant source contribution" (L11-12).

Full Screen / Esc
Printer-friendly Version

Interactive Discussion
Discussion Paper P4942L17 and L24: I am sceptical, whether these references are really suitable to support the rather general statements/explanations in these two sentences. Be sure to avoid referencing secondary sources.
P4942L15-P4947L19: In the introduction, the Unmix and PMF models are described in much detail. The authors might want to check whether such level of detail is really necessary to understand the results and discussion sections of the manuscript (which actually depends on what the latter sections are aiming at).
P4945L16-P4946L2: This section would be more appropriately placed in the experimental section P4947L4-14: This section seems a bit out of place here.

P4947L15-19:
The main goal of the study has to be more clearly defined (see above).
P4947L20-P4948L17: The Materials and methods section is poorly written. It contains paragraphs which have to be moved to other sections (e.g. the first 7 lines are rather introductory than experimental), the information on sampling and analysis is incomplete (I miss details on the time of sampling (in which months/years?), the country the site belongs to (Google maps did help here but it might be worth to mention Montenegro in the manuscript), and data on the manufacturers of the applied instruments and materials), and -most importantly -it does not give the experimental details of the two models applied. These are actually spread among the introduction (see above) and results sections, but they should be brought together here in this section. P4953L14-4955L24: As stated above, the discussion section does not fit into the scopes of AMT as it is focused too much on the source apportionment aspect of the study and hardly at all on the evaluation of the models. C2023